Mammalian cells are the preferred host cells for the production of complex biopharmaceutical proteins as the post-translational modifications are human-compatible both functionally and from a pharmacokinetic point of view. The main relevant cells types are hydridomas, myelomas, CHO (Chinese Hamster Ovary) cells and BHK (Baby Hampster Kidney) cells. The host cells are increasingly cultivated under serum- and protein-free production conditions. The reasons for this are the associated reduction in costs, the reduced interference in the purification of the recombinant protein and the reduction of the potential for introducing pathogens (e.g. prions and viruses). The use of CHO cells as host cells is becoming more and more widespread as these cells adapt to suspension growth in serum- and protein-free medium and moreover are regarded and accepted as safe production cells by the regulatory bodies.
In order to produce a stable mammalian cell line which expresses a heterologous gene of interest, the heterologous gene is generally introduced into the desired cell line together with a selectable marker gene, such as neomycin phosphotransferase (NPT), by transfection. The heterologous gene and the selectable marker gene can be expressed in a host cell, starting from an individual or separate co-transfected vectors. Two to three days after the transfection the transfected cells are transferred into medium containing a selective agent, e.g. G418 when using the neomycin-phosphotransferase gene (NPT gene) and cultivated for a few weeks under these selective conditions. The emergent resistant cells which have integrated the exogenous DNA can be isolated and investigated for the expression of the desired gene product (gene of interest).
For biopharmaceutical production, cell lines with a high stable productivity are required. The expression vectors for production cells are equipped with strong, usually constitutively expressing promoters and enhancers such as CMV enhancer and promoter, for example, to allow high product expression. As the expression of the product has to be guaranteed over the longest possible time, cells are selected which have the product gene stably integrated in their genome. This is done with selectable markers such as e.g. neomycin-phosphotransferase (NPT) and dihydrofolate reductase (DHFR).
By the chance integration of the expression vectors in the host cell genome, cells are obtained with different levels of expression of the desired gene product, as its expression is not determined solely by the strength of the previous promoter or the promoter/enhancer combination. The chromatin structure present at the integration site can affect the level of expression both negatively and positively. Increasingly, therefore, cis-active elements which positively influence the expression at the chromatin level are integrated in expression vectors. These include locus control regions (LCR) which occur for example in the 5′ region of the β-globin genes (Li et al., 2002) and in the 3′ region of the TCRα gene. They cause high tissue-specific expression of a coupled transgene in the chromatin, which is characterised by its independence of position and dependence on copy number. These properties indicate that LCRs are capable of opening chromatin in their native tissue (Ortiz et al., 1997). There are various forms of β-thalassaemia in which the β-globin locus is intact but is not expressed. The reason for the lack of expression is a major deletion in the 5′ direction of the β-globin genes. The deletion of this β-globin LCR leads to a closed chromatin configuration which extends over the entire locus and leads to suppression of gene expression (Li et al., 2002). LCRs colocalise with DNAse I-hypersensitive sites (HS) in the chromatin of expressing cells. The occurrence of HS also indicates open chomatin. The HS contain a series of different general and tissue-specific binding sites for transcription factors. By the interaction of the transcription factors with the DNA the open chromatin structure of HS is produced (Li et al., 2002). Many LCRs are known to be made up of a number of HS the functions of which can be more or less separated from one another. The TCRα gene for example is expressed under endogenous control only in T-cell tissue. The locus exists in various chomatin modes depending on the tissue and expression status. In the 3′ region it has a locus control region which has 8 HS. HS 2-6, a 6 kb partial fragment of the LCR, has a chromatin-opening activity and is not tissue specific. The tissue specificity is imparted to the T-cell-specific expression in the thymus by HS7, 8 and 1 (3 kb). Only in the complete combination of all the HS is the TCRαLCR functionally complete (Ortiz et al., 1997). A more precise subdivision and specification of the individual HS functions of the TCRαLCR can be found in (Ortiz et al., 1999). This Example shows that LCRs are functionally very complex and may be made up of different control elements such as enhancers, silencers and isolators. Other Examples of the division of the LCR functions between various domains are the TCRγ locus and β-globin locus. The former is made up of the DNAse I-hypersensitive site HsA and the enhancer 3′ECγ1. The TCRγ-LCR, in addition to having its usual functions, is also thought to play a part in the recombination of the TCRγ genes (Baker et al., 1999). The β-globin locus has five HS with distinguishable functions which also require the tissue-specific promoter in order to function fully. LCRs could also play another important role in the tissue-specific demethylation of DNA, as DNA methylation results in a closed chromatin structure and the inactivation of genes. A mechanism of activity which activates gene expression by increased histone acetylation would also be possible (Li et al., 2002).
Scaffold/Matrix Attachment Regions (S/MARs) are DNA sequences which bind with a high affinity in vitro to components of the matrix or the scaffold of the cell nucleus. They form the structural and possibly also functional boundaries of chromatin domains (Zahn-Zabal et al., 2001). S/MARs are capable of interacting with enhancers and of locally increasing the accessibility of the DNA in the chromatin and in this way can increase the expression of stably integrated heterologous genes in cell lines, transgenic animals and plants (Klehr et al., 1991; Stief et al., 1989; Jenuwein et al., 1997; Zahn-Zabal et al., 2001). However they cannot totally shield a chromosomal locus from nearby elements in order to allow position-independent expression (Poljak et al., 1994). The effect of the MARs can be used to increase the proportion of (highly) expressing cell clones or transgenic animals in a transfection experiment (McKnight et al., 1992; Zahn-Zabal et al., 2001). However, MARs have also been reported which do not impart high expression but play an important part in the correct regulation of development-specific genes (McKnight et al., 1992).
Isolators are defined as a neutral boundary between neighbouring regions which influence one another, e.g. between active and inactive chromatin (boundary elements). They may restrict the effect of enhancers or isolate entire DNA domains against them and shield stably transfected reporter genes from positional effects (Bell and Felsenfeld, 1999; Udvardy, 1999). Thus, these elements render the expression independent of the genomic position. They may also prevent the silencing of transgenes in the absence of selection pressure (Pikaart et al., 1998). Another presumed function of isolators is the restriction of replication territories (Bell and Felsenfeld, 1999). The first isolators that were described are scs and scs' from Drosophila. They constitute the boundary for the hsp70 heat shock genes and suppress positional effects (Udvardy et al., 1985).
As another element with an isolating function, a GC-rich fragment from the dhfr gene (Chinese Hamster) was found, containing CpG islands (Poljak et al., 1994). The fragment on its own exhibited no influence whatever on reporter gene expression. Situated between an expression promoting SAR and the reporter gene, however, the fragment was able to substantially prevent the expression-enhancing effect of the SAR element. Possibly, this GC-rich fragment blocks the chromatin-opening mechanism of the SAR element and consequently acts as an isolator. Elements with extended CPG islands are methylated with a higher probability as they are recognised by a DNA methyltransferase which converts cytosine into 5-methylcytosine. Consequently, inactive chomatin is formed (Poljak et al., 1994).
Aronow and colleagues defined in the first intron of the human ADA gene (Adenosin deaminase) a new regulatory element which substantially contributes to expression which is dependent on gene copy number but independent of position (Aronow et al., 1995). The element is up to 1 kb in size and only functional when it flanks a 200 bp T-cell-specific enhancer. If only one of the two segments is present or if the segments are wrongly arranged in their sequence and orientation, the element is non-functional as this prevents the formation of DNAse I-hypersensitive sites on the enhancer.
In their Patent WO 02/081677 the firm Cobra Therapeutics describe another chromatin-influencing element. The Ubiquitous Chromatin Opening Elements (UCOEs) are responsible for an open chromatin structure in chromosomal regions with ubiquitously expressed household genes (human hnRNP A2 gene, human β-actin gene, human PDCD2 gene). All these genes have CpG-rich islands in the untranslated regions which are relatively weakly methylated. The absence of methylation of CpG islands indicates that there is active chromatin at this point. The UCOEs help to provide a strength of expression which is independent of the genomic environment and the nature of the cell or tissue.
The firm Immunex also describes cis-active DNA sequences which bring about an increase in expression (U.S. Pat. No. 6,027,915, U.S. Pat. No. 6,309,851). The element referred to as the expression augmenting sequence element (EASE) demands a high expression of recombinant proteins in mammalian cells, is not active in transient expression systems and does not have the typical sequence properties found in LCRs and S/MARs. It is also not a sequence which codes for a trans-activating protein as it does not contain an open reading frame. The fragment is 14.5 kb long, originates from the genomic DNA of CHO cells and can increase the expression of a stably integrated reporter gene eight fold. Over 50% of the activity of the element is restricted to a 1.8 kb long segment, while the first 600 base pairs of this segment are essential for correct function. An additional property of sequence sections with a high EASE activity is the presence of a number of HMG-1(Y) binding sites. HMG-I(Y) proteins belong to the family of the high mobility group non-histone chromatin proteins. They are also referred to as “architechtonic transcription factors” and form a new category of trans-regulators of mammalian genes. HMG-I(Y) proteins recognise 80-rich sequences and bind to their so-called AT hooks (DNA-binding domains) in the small DNA fork. This can lead to local changes in the DNA topology and consequently to altered gene expression.
The authors of U.S. Pat. No. 6,309,841 presume that the effects of EASE are connected with the MTX-induced amplification of the integrated plasmid. In MTX-induced gene amplification, so-called breakage fusion bridge cycles occur. It is easy to imagine a role for the HMG-I(Y) proteins in the structural alteration of the DNA which lead to the formation and removal of the DNA breakages.
Other elements for increasing gene expression in mammalian cells are described in Kwaks et al., 2003. These so-called STAR elements (Stimulatory and Anti-Repressor Elements) originate from the screening of a human gene library with 500 to 2100 bp DNA fragments. The screening was carried out using a specially designed reporter plasmid. The expression of the reporter gene was only possible when it was functionally linked with an anti-repressor element from the human gene bank. With the STAR elements thus obtained the authors were able to protect transgenes from positional effects in the genome of mammalian cells. A comparison with the mouse genome showed that the majority of these STAR elements occur in both the human and the murine genome and are highly conserved within these two species.
A major problem in establishing cell lines with a high expression of the desired protein arises from the random and undirected integration of the recombinant vector in transcription-active or -inactive loci of the host cell genome. As a result, a population of cells is obtained which show completely different expression rates for the heterologous gene, while the productivity of the cells generally follows a normal distribution. In order to identify cell clones which have a very high expression of the heterologous gene of interest it is therefore necessary to check and test a number of clones, resulting in high expenditure of time, labour and costs. Attempts at improving the vector system used for the transfection are therefore directed at even allowing or increasing the transcription of a stably integrated transgene by the use of suitably cis-active elements. The cis-active elements which act at the chromatin level include for example the locus control regions, scaffold-matrix attachment regions, isolators, etc., already described. Some of these elements shield certain genes from the influences of the surrounding chromatin. Others exhibit an enhancer-like activity, although this is restricted to stably integrated constructs. Yet other elements combine several of these functions in themselves. Often it is not clearly possible to assign them precisely to a specific group. In stable cell lines the expression of the transgenic product gene thus underlies chromosomal positional effects to a considerable extent. This phenomenon is based on the influence of the chromatin structure and/or the presence of intrinsic regulatory elements at the integration site of the foreign DNA. This leads to very variable expression levels. During the selection of cells, therefore, frequently clones with a very low or completely absent product expression are frequently produced. These chromosomal positional effects are also the reason why the generation of stable production cell lines which express a high level of a therapeutic protein is generally a time consuming, high-capacity and expensive process. Stable cell lines with high productivity are usually produced by selection with positive selectable markers, frequently combined with agent-induced gene amplification (e.g. dihydrofolate reductase/methotrexate or glutamine synthetase/methioninesulfoximine). The pools and clones which are produced with this selection strategy are investigated for high and stable expression in a complex screening process. The majority of the clones produce no or only average amounts of product and only a few are high producers. The proportion of high producers in a mixed population can be increased, for example, by a mutation in the selectable marker (Sautter and Enenkel, 2005, WO 2004/050884). However, it is desirable to further increase the specific productivity of each individual clone as well as the proportion of high producers within a transfected cell population.
The specific productivity of stably transfected cells, particularly CHO- or other production-relevant cells, and the proportion of high producers in a transfection batch should be increased. This should result in the last analysis in a more efficient cell line development. Consequently more and higher producing cell lines could be established in a shorter time and thus save on labour, time and costs.